PHASE 3.22 — LAYMAN RESULTS EXPLANATION
MULTIPLIER PROGRAMME BASELINE — 1x WAGER ONLY

WELCOME TO THE NEXT RESEARCH LAYER
Phase 3.22 begins the multiplier / wager programme.

But we deliberately do NOT change the wager yet.

The purpose of 3.22 is to measure exactly what the frozen decision policy does under the existing 1-unit wager before we start experimenting with larger or conditional stakes.

This gives every future multiplier experiment a clean reference point.

WHAT IS FROZEN?
The decision policy is the Phase 3.20 checkpoint validated by Phase 3.21.

HIT / STAND / DOUBLE / SPLIT logic is now protected.

Future multiplier experiments may change how much is wagered, but they must not quietly change the decision policy at the same time. That separation lets us tell whether a future improvement or failure came from wagering rather than from another blackjack decision correction.

WHAT CHANGED IN 3.22?
Decision rules: NOTHING.
Wager: NOTHING — still 1 unit.
Shoe/rules/information boundary: NOTHING.

What changed is the instrumentation.

For the first time the multiplier programme prospectively records the bankroll trajectory after every hand, each session's peak bankroll, maximum drawdown, whether multiplier thresholds were reached at ANY TIME during a session, and survival.

BASELINE ENDING BANKROLL
Primary average ending bankroll: 98.412.
Holdout: 98.496.

These exactly preserve the frozen-policy result.

THE 30-HAND TRAJECTORY
The average bankroll declines gradually rather than collapsing suddenly.

Primary:
Hand 1  = 99.901
Hand 10 = 99.427
Hand 20 = 98.930
Hand 30 = 98.412

Holdout:
Hand 1  = 99.933
Hand 10 = 99.495
Hand 20 = 98.975
Hand 30 = 98.496

This trajectory is extremely useful for multiplier research. We can later see not merely whether a multiplier finishes higher or lower, but WHEN its effect begins to appear.

PEAK BANKROLL — NOW MEASURED PROSPECTIVELY
Highest bankroll observed anywhere:
Primary: 126.0 at Hand 29.
Holdout: 125.5 at Hand 29.

Mean per-session peak:
Primary: 104.304.
Holdout: 104.317.

So a typical session does experience some upside above the 100-unit start even though the overall average finishes slightly below 100.

833 Primary sessions and 889 Holdout sessions never exceeded the original 100-unit bankroll at any point.

DRAWDOWN
Mean per-session maximum drawdown from the running peak:
Primary: 8.317 units.
Holdout: 8.256 units.

This becomes one of our important controls.

A future multiplier that increases the ending bankroll but produces an enormous drawdown is not automatically a better policy. We will be able to measure reward and bankroll stress together.

DID ANY SESSION REACH THE MULTIPLIERS DURING PLAY?
No.

Under the 1x baseline:
Ever reached 2x (200 units): 0.00%.
Ever reached 3x (300): 0.00%.
Ever reached 5x (500): 0.00%.
Ever reached 10x (1,000): 0.00%.

This is stronger information than our earlier end-of-session statistic because we are now prospectively checking the bankroll after every hand.

Therefore the old question is resolved cleanly:
with 30 hands and a flat 1-unit wager, these multiplier targets are not merely failing to be retained at Hand 30 — they are not being reached during the session at all in these cohorts.

SURVIVAL
Sessions ever reaching bankroll <=0:
Primary: 0.00%.
Holdout: 0.00%.

That is expected with a 100-unit bankroll and tiny flat wager, but it establishes the survival baseline against which more aggressive wagering can be judged.

WHY THIS BASELINE IS IMPORTANT
Suppose a future wager architecture reaches 2x in some sessions.

We will now be able to ask:
- How often?
- At what hand?
- Was the gain retained?
- What happened to average ending bankroll?
- What happened to drawdown?
- How many sessions went broke?
- Did Primary and Holdout replicate?
- Was 3x/5x/10x achieved only through freak routes or with meaningful frequency?

This prevents us from celebrating a spectacular isolated bankroll while ignoring what happened to the rest of the population.

WHAT SHOULD 3.23 DO?
Now we can introduce the FIRST genuine wager experiment.

The safest scientifically useful first step is not 10x wagering immediately.

Phase 3.23 should test a simple constant 2x base wager:
- same frozen decision policy;
- same 100-unit starting bankroll;
- same 30 hands;
- wager scale changed from 1 unit to 2 units only.

This gives us the cleanest possible answer to:
'What happens when we double exposure while changing absolutely nothing else?'

We should measure the full 3.22 metric set again.

If constant 2x behaves as expected, later experiments can investigate 3x, 5x, 10x or conditional/dynamic multiplier structures in controlled steps.

IMPORTANT DISTINCTION
The target '2x bankroll' means reaching 200 bankroll units.

A '2x wager' in Phase 3.23 would mean doubling the base wager from 1 unit to 2 units.

They are not the same thing.

PUBLICATION NOTE
This Layman Results report contains the complete explanatory interpretation for Phase 3.22.
